Towards Energy Efficient Memristor-based TCAM for Match-Action Processing

Saleh, Saad; Goossens, Anouk S.; Banerjee, T.; Koldehofe, Boris

doi:10.1109/igsc55832.2022.9969354

Cited by 6 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The output of dAQM is processed by a set of egress packet processing stages until ready for processing by the egress packet queues. dAQM can be implemented in four possible methods; (1) Programming dAQM inside the Match-Action tables [37], (2) Using the emerging matchaction frameworks that support more expressive line rate network functions [38]- [46], (3) Network Function Virtualization (NFV)-based software implementation, (4) Programming dAQM inside an SDN-based controller.…”

Section: Proposed Daqm Algorithmmentioning

confidence: 99%

Adaptive In-Network Queue Management using Derivatives of Sojourn Time and Buffer Size

Saleh,

Shu,

Koldehofe

2024

NOMS 2024-2024 IEEE Network Operations and Management Symposium

Self Cite

View full text Add to dashboard Cite

Active Queue Management (AQM) algorithms are heavily used in packet processors to maintain an optimal queue size and avoid issues like Bufferbloat. Despite the remarkable performance, the traditional AQM algorithms face a major challenge of estimating the accurate queue congestion due to bursty network conditions. The major reason is the use of baseline queue statistics for congestion estimation like delay and sojourn time for Random Early Detection (RED) and Controlled Delay (CoDel), respectively. In this paper, we propose a novel dAQM algorithm that uses advanced traffic statistics like three higher-order derivatives of sojourn time and buffer size along with the baseline sojourn time and buffer size for accurate congestion estimation. dAQM adjusts its drop rate based on the continuously varying congestion to cater to the needs of bursty traffic. We simulated dAQM in ns-3 and analyzed its performance for FTP traffic by variation in traffic load and packet sizes. The results showed that dAQM provides at least 25% and 39.7% reduction in packet loss ratio and flow completion time, respectively, as compared to the traditional AQM algorithms.

show abstract

Section: Proposed Daqm Algorithmmentioning

confidence: 99%

Adaptive In-Network Queue Management using Derivatives of Sojourn Time and Buffer Size

Saleh,

Shu,

Koldehofe

2024

NOMS 2024-2024 IEEE Network Operations and Management Symposium

Self Cite

View full text Add to dashboard Cite

show abstract

“…Building on our previous work [6] focusing on understanding the challenges in design of such general purpose architectures, our contributions are threefold, (1) development of a memristor-based TCAM architecture, called TCAmM CogniGron , for match-action processors, (2) analysis of power, latency and resource density of the proposed architecture on Nb-doped SrTiO 3 memristor, (3) understanding of the trade-offs of configuring TCAmM CogniGron for power-efficient computing.…”

Section: Introductionmentioning

confidence: 99%

TCAmM^CogniGron: Energy Efficient Memristor-Based TCAM for Match-Action Processing

Saleh

Goossens

Banerjee

et al. 2022

2022 IEEE International Conference on Rebooting Computing (ICRC)

Self Cite

View full text Add to dashboard Cite

The Internet relies heavily on programmable matchaction processors for matching network packets against locally available network rules and taking actions, such as forwarding and modification of network packets. This match-action process must be performed at high speed, i.e., commonly within one clock cycle, using a specialized memory unit called Ternary Content Addressable Memory (TCAM). Building on transistorbased CMOS designs, state-of-the-art TCAM architectures have high energy consumption and lack resilient designs for incorporating novel technologies for performing appropriate actions. In this article, we motivate the use of a novel fundamental component, the 'Memristor', for the development of TCAM architecture for match-action processing. Memristors can provide energy efficiency, non-volatility and better resource density as compared to transistors. We have proposed a novel memristorbased TCAM architecture called TCAmM CogniGron , built upon the voltage divider principle and requiring only two memristors and five transistors for storage and search operations compared to sixteen transistors in the traditional TCAM architecture. We analyzed its performance over an experimental data set of Nbdoped SrTiO 3 -based memristor. The analysis of TCAmM CogniGron showed promising power consumption statistics of 16 µW and 1 µW for match and mismatch operations along with twice the improvement in resources density as compared to the traditional architectures.

show abstract

Analog In-Network Computing through Memristor-based Match-Compute Processing

Saleh,

Goossens,

Shu

et al. 2024

IEEE INFOCOM 2024 - IEEE Conference on Computer Communications

View full text Add to dashboard Cite

Current network functions consume a significant amount of energy and lack the capacity to support more expressive learning models like neuromorphic functions. The major reason is the underlying transistor-based components that require continuous energy-intensive data movements between the storage and computational units. In this research, we propose the use of a novel component, called Memristor, which can colocalize computation and storage, and provide computational capabilities. Building on memristors, we propose the concept of match-compute processing for supporting energy efficient network functions. Considering the analog processing of memristors, we propose a Probabilistic Content Addressable Memory (pCAM) abstraction which can provide analog match functions. pCAM provides deterministic and probabilistic outputs depending upon the closeness of match of an incoming query with the specified network policy. pCAM uses a crossbar array for line rate matrix multiplications on the match outputs. We proposed a match-compute packet processing architecture and developed the programming abstractions for a baseline network function, i.e., Active Queue Management, which drops packets based upon the higher-order derivatives of sojourn times and buffer sizes. The analysis of match-compute processing over a physically fabricated memristor chip showed only 0.01 fJ/bit/cell of energy consumption, which is 50 times less than the traditional match-action processing.

show abstract

Towards Energy Efficient Memristor-based TCAM for Match-Action Processing

Cited by 6 publications

References 17 publications

Adaptive In-Network Queue Management using Derivatives of Sojourn Time and Buffer Size

Adaptive In-Network Queue Management using Derivatives of Sojourn Time and Buffer Size

TCAmM^CogniGron: Energy Efficient Memristor-Based TCAM for Match-Action Processing

Analog In-Network Computing through Memristor-based Match-Compute Processing

Contact Info

Product

Resources

About

Towards Energy Efficient Memristor-based TCAM for Match-Action Processing

Cited by 6 publications

References 17 publications

Adaptive In-Network Queue Management using Derivatives of Sojourn Time and Buffer Size

Adaptive In-Network Queue Management using Derivatives of Sojourn Time and Buffer Size

TCAmMCogniGron: Energy Efficient Memristor-Based TCAM for Match-Action Processing

Analog In-Network Computing through Memristor-based Match-Compute Processing

Contact Info

Product

Resources

About

TCAmM^CogniGron: Energy Efficient Memristor-Based TCAM for Match-Action Processing